BACKGROUND & AIMS: Neoadjuvant chemotherapy prior to liver surgery for colorectal metastases can cause marked steatosis (≥ 33%) and steatohepatitis defined by non-alcoholic fatty liver disease activity score (NAS) as adverse effects on liver parenchyma. The aim of this study was to evaluate the steatosis level prior to liver resection using proton magnetic resonance spectroscopy ((1)H MRS) and to compare it with digital quantification of steatosis (DQS) and "classical" histopathology. METHODS: (1)H MRS at 3T evaluated steatosis in 35 patients with colorectal liver metastasis, planned for liver resection. Non-tumorous liver parenchyma samples were obtained after surgery for classical histopathology and DQS utilising automated software for quantification of histopathological slides using image processing. RESULTS: Classical histopathology defined marked steatosis in nine patients. Histopathology was less reliable than DQS (interclass correlation coefficient - ICC 0.771) or (1)H MRS (ICC 0.722) in steatosis estimation. (1)H MRS showed very similar steatosis levels and high reliability compared to DQS (ICC 0.955). Steatohepatitis was observed in seven patients (NAS ≥ 4) and (1)H MRS was able to predict it with 100% sensitivity and 89% specificity at threshold 10.9%, without knowing lobular inflammation or hepatocyte ballooning. BMI was significantly higher in the groups with marked steatosis and steatohepatitis. Standard blood tests or chemotherapy had no predictive value. CONCLUSIONS: (1)H MRS is a reliable non-invasive tool for steatosis assessment, and interestingly, it was able to predict steatohepatitis defined by NAS ≥ 4 in patients planned for liver resection of colorectal metastases after neoadjuvant chemotherapy.
BACKGROUND & AIMS: Neoadjuvant chemotherapy prior to liver surgery for colorectal metastases can cause marked steatosis (≥ 33%) and steatohepatitis defined by non-alcoholic fatty liver disease activity score (NAS) as adverse effects on liver parenchyma. The aim of this study was to evaluate the steatosis level prior to liver resection using proton magnetic resonance spectroscopy ((1)H MRS) and to compare it with digital quantification of steatosis (DQS) and "classical" histopathology. METHODS: (1)H MRS at 3T evaluated steatosis in 35 patients with colorectal liver metastasis, planned for liver resection. Non-tumorous liver parenchyma samples were obtained after surgery for classical histopathology and DQS utilising automated software for quantification of histopathological slides using image processing. RESULTS: Classical histopathology defined marked steatosis in nine patients. Histopathology was less reliable than DQS (interclass correlation coefficient - ICC 0.771) or (1)H MRS (ICC 0.722) in steatosis estimation. (1)H MRS showed very similar steatosis levels and high reliability compared to DQS (ICC 0.955). Steatohepatitis was observed in seven patients (NAS ≥ 4) and (1)H MRS was able to predict it with 100% sensitivity and 89% specificity at threshold 10.9%, without knowing lobular inflammation or hepatocyte ballooning. BMI was significantly higher in the groups with marked steatosis and steatohepatitis. Standard blood tests or chemotherapy had no predictive value. CONCLUSIONS: (1)H MRS is a reliable non-invasive tool for steatosis assessment, and interestingly, it was able to predict steatohepatitis defined by NAS ≥ 4 in patients planned for liver resection of colorectal metastases after neoadjuvant chemotherapy.
Authors: Jing Qi; Yuman Fong; Leonard Saltz; Michael I D'Angelica; Nancy E Kemeny; Mithat Gonen; Jinru Shia; Amita Shukla-Dave; William R Jarnagin; William M Jarnagin; Richard K G Do; Lawrence H Schwartz; Jason A Koutcher; Kristen L Zakian Journal: NMR Biomed Date: 2012-09-07 Impact factor: 4.044